5 ms·
> The languages do seem to be responsible for at least 1 order of magnitude. Compared to what? C? I'm not talking about C, but about any modern language. > D
by pron 10y ago
> The languages do seem to be responsible for at least 1 order of magnitude.
Compared to what? C? I'm not talking about C, but about any modern language.
> Do they have to, though?
Yes. Or, at least, everybody (including me when I first saw them) says "this cannot possibly be this complicated", and after understanding them, everybody says, "oh, OK". But let me put it another way: if the modern world really only requires simple software, then our work is nearly done. Some would use Python, some would use Java, some would use Haskell -- but if the largest software system needs to be ~100KLOC, then none of it matters too much. Writing such a piece of software isn't hard regardless of what language you use (or, rather, the difficulty is in the essential complexity; there's not much accidental complexity in 100KLOC). Even assuming one methodology would be 15% than another, it wouldn't make much difference to the bottom line because producing 100KLOC software is cheap anyway. And if that were the case, investing in programming languages would be even a bigger waste, as investing in simplifying specifications would have a much bigger impact, and would cost a lot less.
But just to get a sense, GHC -- which is a compiler, and, as you've noted, compilers tend to be small -- is about 400KLOCs of Haskell. The Linux kernel is over 15MLOC of C. Reduce that by an order of magnitude, you still get 1.5MLOC, and that's just for an OS kernel.
> So far, you're the only one I met that challenges that perception.
I don't argue that there isn't a lot of waste. I argue that even without all that waste, we'd still need software that's very large (or that, alternatively, waste is unavoidable). There are no signs that Haskell reduces this waste at all, or that it dramatically reduces the size of programs.
- loup-vaillant 10y ago> Compared to what? C? I'm not talking about C, but about any modern language. The domain they tackle is personal computing, which means Kernel, Windowing system, remote communication (web/mail), multimedia… So, yeah: mostly C and C++, by the look of currently popular programs. That said, much of the (apparently) needlessly complex stuff I have seen was written in C++, and it did look like they didn't have the real-time requirements or resources constraints that would justify the use of such a monster of a language. > Or, at least, everybody (including me when I first saw them) says "this cannot possibly be this complicated", and after understanding them, everybody says, "oh, OK". I have yet to reach the second stage. And on one occasion, I did reach a reasonable understanding of the whole system. It definitely had to be very complex to match the specification, but the specifications themselves didn't match the end user's needs. > (or that, alternatively, waste is unavoidable) That's the alternative I'm most scared of. I cannot comprehend unavoidable waste, but I can't rule it out either. Alternatively, I've came across the idea of not solving some problems¹, because the return on investment is just crap. Okay, when safety is involved, you probably cannot do that. Still, the idea that the 80/20 rule is sometimes more like 99.9/0.1 is enticing. Sometimes, full automation is not best. For instance last I checked, the best Chess player ever is a human-computer team, not a computer. [1] Stop Over-Engineering https://www.youtube.com/watch?v=GRr4xeMn1uU https://www.youtube.com/watch?v=GRr4xeMn1uU
- pron 10y ago> I've came across the idea of not solving some problems That, too, cannot be solved by a programming language :) But much of complex software deals with things that could not possibly have been solved more efficiently (or even efficiently enough) by humans (like sensor fusion, package tracking, manufacturing control etc.). Also, we're pretty far from full automation. Nobody trusts computers to make the decisions in air-traffic control systems, power plant control software, or even in ERP systems. The reason we keep building large software is that -- in spite of many problems -- they really do work (in the sense of achieving their goal of higher-throughput; whether or not a higher throughput of flights or of business deals is good or bad for humanity is an entirely separate question).
- tome 10y ago> if the largest software system needs to be ~100KLOC, then none of it matters too much. Writing such a piece of software isn't hard regardless of what language you use I think this is absolutely the difference between you and proponents of Haskell. Personally I have found working on 100k LOC codebases very hard in Python and easy in Haskell.
- pron 10y agoI wouldn't know because I've never written anything in Python. How many 100KLOC non-compiler programs would say have been written in Haskell? It would be great if the team behind one of them would write a technical report so we'd at least slowly get to find out if it's actually easier to work with Haskell than with Python, and if so, by how much.
- tome 10y agoI agree that would be great. I can't help thinking that you're holding Haskell up to a standard to which you do not hold Rust, Go or Julia. Have you asked their proponents for technical reports on Hacker News threads?
- pron 10y agoIt is Haskell that holds itself to a higher standard. Rust, Julia and Go don't make claims that are anywhere as extreme as Haskell's fans (e.g., see on this page the claim that Haskell's high abstractions -- optical profunctors or whatever -- are "load-bearing materials" to other languages' "mud". Rust has presented itself as a safe alternative to C/C++; Go presents itself as a high-performance language with simple concurency (or Java without the JVM); Julia presents itself as a high-performance alternative to Matlab, R or NumPy. But Haskell has sort of painted itself into a corner. Because the approach is so foreign and the learning-curve so steep -- i.e., the adoption cost is high -- if it didn't make outlandish claims (like, if it compiles it works) then no one in industry would consider using it. I don't need to know by how much Go lowers development costs because it makes no claims that it does, so I simply assume that it doesn't. That it is faster than Python, easy to learn in a day or two, and that it compiles down to a native executable -- are all trivial to verify. If you want, all of its claims are supported by plentiful data. But other than that: yes! It is trivial to verify that Rust indeed fulfills its claims, but it is not trivial to verify that that's enough, namely, that overall, the overall cost of developing in Rust is lower than in C++, which is why I wouldn't switch from C++ to Rust without seeing that at least the costs are comparable. I'm also eagerly waiting for data about Rust's concurrency approach. I would also tell you this: while I find the pure-functional aesthetically unattractive for interactive programs and am very much impressed by Clojure's and Erlang's approaches to state, and while I've personally written a semi-popular Clojure library, I have expressed skepticism about Clojure's suitability for large-scale projects. How well a language works in big software is something that you simply cannot extrapolate from experience in small projects, and as much as I like Clojure, I have serious doubts about its applicability, and I would not use it to write a large system without the kind of data I expect from Haskell. But all those languages (Erlang and Clojure included) have social advantages that make this data much more easy to come by: they are used by people who aren't enamored with the language itself, who aren't PL enthusiasts, and are much more goal-oriented. Their judgment stems mostly from how well things work, not how interesting the language is. So articles with pertinent data (maybe not enough to risk a large project, but certainly more than those you find for Haskell) are more abundant. So I don't need to beg for technical reports as much because 1. no wild claims that aren't supported by data are made, and 2. there's data out there (for Go, Erlang, Clojure) or it's coming (Rust).